Board Attributes, Political Connections and External Auditor Pricing Behaviour in Nigeria: Statistical versus Machine Learning Analyses

Anchored on Simunic’s audit fees model, this study examined the influence of a few board attributes on pricing behavior of external auditors in Nigeria. Besides, political connections that are largely prevalent among corporate entities in Nigeria, is also introduced as part of explanatory variables. Data were gathered from audited annual accounts and reports of 50 quoted non-financial firms for 7 years from 2014 – 2020, culminating into 350 data/set observations. Both statistical learning (SL) and machine learning (ML) data analytic techniques were adopted to estimate model specified for the study. The results indicated that firm size and political connections are effective determinants of pricing behaviour of auditors, accounting for 80% variations therein. This is similar to results obtained from machine language analysis. However, board attributes, as well as profitability (return on total assets), in addition to firm size and political connections, displayed positive and statistical significant relationship when BIG4 was adopted as proxy for pricing behavior of auditors. The statistical and machine learning analyses accounted for 21% and 86% variations in pricing behavior of auditors respectively. This study largely conforms to extant studies and aligns with demand-based perspective linking board attributes and audit pricing behavior. Consequently, board attributes, political connections, firm size and profitability are active variables available to firms, which must be effectively managed in order to influence the amount charged by external auditors in Nigeria.